A Style-Based Generator Architecture for Generative Adversarial Networks

Tero Karras, Samuli Laine, Timo Aila · CVPR / arXiv · 2019 · art-history · record karras2019-stylegan
tag layer · index · concepts · authors · עברית

Where this record's text came from. faceted 2026-08 by reading the locally held text (pdftext); every evidence string verified as an exact substring of that text before merging. Replication_status left 'untested': the paper reports that parallel work corroborates the general superiority of style-based design ("corroborating the large-scale ImageNet measurements made in parallel work"), but no one has re-run this paper's own measurements. The third limitation quote is a fragment because the PDF extraction interleaved a figure caption mid-sentence; it is reproduced exactly as it appears in the article text. The paper also declines to add an encoder network for existing disentanglement metrics, which is a scope decision rather than a stated limitation, so it is not listed. 2026-08-14, 2026-08-16 and 2026-08-25 schema-0.3 passes: every quote verified as an exact substring of the staged text and anchored at merge time; concept terms verified present; every edge endpoint an existing node. Update, 15 September 2026: EXCEPTION, measured with the merge gate's own predicate (Paper.locate): 1 quote(s) on this record are not found as one exact string in the current staged text - content_tags.statistics[0].evidence. The sentence above does not hold for them as written. Where a row records why - the quote is grounded in another copy of the paper, joins two passages with an ellipsis, or differs from the staged text by extraction noise - that note stays on the row.

Some words in these quotes contain the wrong character. The PDF this text was extracted from substituted 2 characters inside a word — a currency sign or a stray accent where a letter pair belongs, as in signi¢cantly for significantly, or a single ligature glyph, as in conﬁdence. They are valid characters in the wrong place, so they survive every encoding check. The quotes keep them for the same reason as above, and a search of this page will not find the affected words.

1 quote on this page was cut from a different copy of this article. This project holds two extractions of this paper: the plain text every other quote here is anchored in, and the publisher’s own HTML edition, which we are not licensed to republish. It is verbatim in the second and cannot be found in the first — not because the tagging is wrong, but because the extraction damaged the passage: a page number dropped into the middle of a number, a formula moved out of its sentence, a ligature where two letters belong. The record names the copy for each one. The quote is the paper’s; the extraction is not.

How to read this page. The article itself is not reproduced here — this paper is not cleared for redistribution, so what follows is the tag layer alone. The evidence sentences are withheld with the article. The table at the foot lists every tag the record carries; each row names the facet it belongs to, and the whose words column says whether the tag records the paper's own wording or the tagger's sentence about the paper. Every tag links to each other paper in the corpus carrying it. The machine-readable edition of the same thing is the TEI file beside this one, where the tags live in standOff, and the article is at its publisher via the link below. When we have changed our mind. Some entries in the table begin with a word in capitals and the reason for it. EVIDENCE CORRECTED — the finding still stands, but we had quoted the wrong sentence from the paper, and the quotation has been replaced with the right one. RETRACTED — the entry itself was wrong; it stays on the page, marked, with the reason, rather than being deleted. WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below. Each mark carries its own reason, in brackets, because the reasons differ from one entry to the next. An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason. A line reading Audit and a date is a later re-reading of the entry against its whole sentence: DOUBTFUL means it was kept, and the reason says what is loose about it. Quotation marks inside the tagger's sentence. Where a tag is the tagger's own sentence, a phrase in quotation marks inside it is borrowed from the paper — the tagger's wording around it is not. The paper's own sentence is the evidence quote for that row; on this page it is withheld with the article, and the article is at its publisher. Nothing inside quotation marks anywhere on this page is ours: if it is quoted, it is the paper's. Whose words a concept carries. A concept entry records a term the paper itself uses, so the term, the sense given for it, and its evidence are the paper's own words, quoted where the paper fixes them. This is the opposite of a sense entry in the tag table, which is the tagger's sentence about how the paper uses a term. Where a concept entry could not be grounded in the paper's words, the entry says so on its own row.

Other fields define these differently the same word, defined another way in another discipline
deep learning 11 fields · standard-deviation 3 fields

Tags on this paper

Every tag the record carries, each linking to every other paper in the corpus that carries it. A tag appears here whether or not its words occur in the article — the spans below are only the ones a sentence in this paper licenses. Of the 10 concept tags on this paper, 6 are grounded in a quoted sentence.

generative adversarial network image synthesis style transfer deep learning style spatially invariant statistics disentanglement perceptual path length linear separability coarse spatial resolutions each style controls exactly one convolution: AdaIN normalises the channel to zero mean and unit variance first, so the incoming style's scale and bias do not depend on the previous statistics and are overridden by the next AdaIN style localisation is trained for, not only architected: a percentage of images are generated from two latent codes with a crossover point, which stops the network assuming adjacent styles are correlated stochastic detail is supplied by the architecture: per-pixel noise is added after each convolution the division of labour between style and noise is enforced by the discriminator rather than by design: using noise for a global property would produce spatially inconsistent decisions and be penalised the intermediate latent space is defined by what it is NOT required to do - it need not follow the training data's density, which is the constraint the paper blames for entanglement in the input space the path length is defined as a limit under infinitely fine subdivision and computed as an approximation with a small finite epsilon instead separability is measured without an encoder: auxiliary attribute classifiers label generated images, the least confident half is discarded, and a linear SVM is fitted per attribute on the latent-space points generator latent disentangle images synthesis interpolation network AdaIN adaptive instance normalization (AdaIN) FFHQ FID mean average of standard-deviation Standard deviation system-or-tool-building

Legend

relation to other work	`claim_relations`	how this paper stands to another: agrees, contradicts, extends, qualifies
claim type	`claim_type`	the shape of the claim: existence, mechanism, method, critique, prediction, null
concept edge	`concept_edges`	an asserted relation between two concepts, with the sentence asserting it
concept	`concepts`	an idea the paper works with, quoted where the paper uses it
content tag	`content_tags`	a finding, definition, quantity, statistic or variable named in the paper
control	`controls`	an experiment run to eliminate a rival explanation
limitation	`limitations`	what the paper says its evidence cannot support
mechanism	`mechanisms`	how the reported effect is said to come about
method	`method`	how the paper produced its evidence
phenomenon	`phenomenon`	the effect or regularity under study
sense	`senses`	a distinct use of a term, quoted where the paper fixes it
transfer	`transfers`	a method or concept carried in from another field

The paper

The full text is not reproduced here. This article is not cleared for redistribution, so the page carries the tag layer only. Its evidence sentences are withheld; the grounding status remains with each tag, and the article itself is at https://arxiv.org/abs/1812.04948. The reading view exists for the papers whose licence permits it.

Every tag on this page

claim_relations 4 · claim_type 1 · concept_edges 1 · concepts 6 · content_tags 9 · controls 1 · limitations 9 · mechanisms 7 · method 1 · phenomenon 1 · senses 5 · transfers 1

Relations on this page carry a quotation for the claim. Where a relation was separately evidenced — and so tested against the vocabulary for its type — the row says so; none of 1 relation here carry that mark. The rest are the tagger’s reading of the sentence quoted beside them.

facet	tag	field	conf.	whose words
claim_type	method Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_type`		the paper's words
phenomenon	latent-space disentanglement in GAN image synthesis Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`phenomenon`		the paper's words
method	system-or-tool-building Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: system-or-tool-building	`method`		the paper's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: each style controls exactly one convolution: AdaIN normalises the channel to zero mean and unit variance first, so the incoming style's scale and bias do not depend on the previous statistics and are overridden by the next AdaIN	`mechanisms[0]`	0.9	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: style localisation is trained for, not only architected: a percentage of images are generated from two latent codes with a crossover point, which stops the network assuming adjacent styles are correlated	`mechanisms[1]`	0.9	the tagger's words
mechanisms	the paper's reason - that a traditional generator must otherwise invent spatially-varying pseudorandom numbers from earlier activations - is the preceding paragraph, not this span Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: stochastic detail is supplied by the architecture: per-pixel noise is added after each convolution	`mechanisms[2]`	0.9	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the division of labour between style and noise is enforced by the discriminator rather than by design: using noise for a global property would produce spatially inconsistent decisions and be penalised	`mechanisms[3]`	0.9	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the intermediate latent space is defined by what it is NOT required to do - it need not follow the training data's density, which is the constraint the paper blames for entanglement in the input space	`mechanisms[4]`	0.9	the tagger's words
mechanisms	the second adjustment - dividing by epsilon squared because the perceptual distance is quadratic - is a separate sentence far down the section; a span reaching it crosses a footnote, Table 3 and a figure caption Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the path length is defined as a limit under infinitely fine subdivision and computed as an approximation with a small finite epsilon instead	`mechanisms[5]`	0.9	the tagger's words
mechanisms	the conditional entropy and its exponentiation into the final score are described in the sentences after this span Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: separability is measured without an encoder: auxiliary attribute classifiers label generated images, the least confident half is discarded, and a linear SVM is fitted per attribute on the latent-space points	`mechanisms[6]`	0.9	the tagger's words
limitations	the input latent space, constrained to follow the training-data density, has some unavoidable entanglement Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: any disentanglement claim about the input latent space: THE INPUT LATENT SPACE MUST FOLLOW THE PROBABILITY DENSITY OF THE TRAINING DATA, AND WE ARGUE THAT THIS LEADS TO SOME DEGREE OF UNAVOIDABLE ENTANGLEMENT	`limitations[0]`		the tagger's words
limitations	The perceptual path length measurement is admitted to be biased in favour of the input latent space Z. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the perceptual path length comparison, biased against the paper's own conclusion: THIS MEASUREMENT IS IN FACT SLIGHTLY BIASED IN FAVOR OF THE INPUT LATENT SPACE Z	`limitations[1]`		the tagger's words
limitations	Continued FID improvement during training comes at the cost of a more entangled representation (rising path lengths). Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: any use of FID as the training objective: THE SLOWLY RISING PATH LENGTHS INDICATE THAT THE IMPROVEMENTS IN FID COME AT THE COST OF A MORE ENTANGLED REPRESENTATION	`limitations[2]`		the tagger's words
limitations	The new FFHQ face dataset was crawled from Flickr and therefore inherits that site's biases. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: every result and every model trained on FFHQ: the images were crawled from Flickr, THUS INHERITING ALL THE BIASES OF THAT WEBSITE, then automatically aligned and cropped	`limitations[3]`		the tagger's words
limitations	The central term has no agreed definition and the paper proceeds on a shared goal instead. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: any disentanglement claim made here: the paper states there are various definitions of disentanglement and adopts a common goal - a latent space of linear subspaces each controlling one factor of variation - rather than one of them	`limitations[4]`	1.0	the tagger's words
limitations	The two metrics reported are new because the existing ones cannot be run here, so the numbers compare to nobody. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the disentanglement figures given here: the authors state the previously proposed metrics require an encoder network mapping images to latent codes, which this architecture does not have, so the two metrics used are new ones of their own	`limitations[5]`	1.0	the tagger's words
limitations	The perceptual distance the headline metric is built on is another network's embedding, fitted to human ratings. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the perceptual path length reported here: it is built on a pairwise image distance computed as a weighted difference between two VGG16 embeddings, with the weights fitted so the metric agrees with human perceptual similarity judgments	`limitations[6]`	1.0	the tagger's words
limitations	The mechanism claimed for the improvement is posited rather than shown. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the explanation offered for why this generator disentangles: the authors posit that there is pressure on the generator because generating realistic images should be easier from a disentangled representation than from an entangled one	`limitations[7]`	1.0	the tagger's words
limitations	What the paper controls under the name 'style' includes the pose and the shape of the face. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: any reading of this model's 'style' as a manner of depiction: copying the coarse-resolution styles transfers pose, general hair style, face shape and eyeglasses, which are properties of the subject rather than of the manner	`limitations[8]`	1.0	the tagger's words
transfers	the style-transfer idea of separating content from style, taken from that literature and built into the generator architecture so that style is injected at each convolution layer - the network is named after it Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`transfers[0]`	0.85	the tagger's words
controls	an ablation table beginning from a Progressive GAN baseline, with each row adding one change to the row above Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: attributing the improvement to the architecture as a whole - a table where every row differs from the previous one by a single change is what makes any single change attributable · outcome (stated by the tagger, no quotation): held	`controls[0]`	0.85	the tagger's words
concepts	style Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: style	`concepts[4]`		the paper's words
concepts	spatially invariant statistics Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: spatially invariant statistics	`concepts[5]`		the paper's words
concepts	disentanglement Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: disentanglement	`concepts[6]`		the paper's words
concepts	perceptual path length Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: perceptual path length	`concepts[7]`		the paper's words
concepts	linear separability Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: linear separability	`concepts[8]`		the paper's words
concepts	coarse spatial resolutions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: coarse spatial resolutions	`concepts[9]`		the paper's words
senses	'style' - REUSED BY THE AUTHORS' OWN DECISION, FOR ARCHITECTURAL RESEMBLANCE: they say plainly that they choose to reuse the word for the vector y because similar network architectures are already used in feedforward style transfer and image-to-image translation, so the word marks a family of architectures rather than a property of a picture Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[0]`	1.0	the tagger's words
senses	style, once more removed - AS SPATIALLY INVARIANT STATISTICS: the paper's warrant is the style transfer literature's finding that spatially invariant statistics encode style while spatially varying features encode the instance, which is the same move gatys2015 made with a Gram matrix Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[1]`	1.0	the tagger's words
senses	disentanglement - A GOAL STANDING IN FOR A DEFINITION: the paper records that the literature holds various definitions and adopts the shared aim of a latent space of linear subspaces, then measures that aim with two metrics of its own devising Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[2]`	1.0	the tagger's words
senses	style AT COARSE RESOLUTION - POSE, FACE SHAPE AND EYEGLASSES: the styles copied from the coarse spatial resolutions carry high-level aspects of the subject itself, so at this level the word ranges over what the picture is OF and not only how it looks Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[3]`	1.0	the tagger's words
senses	style ON NON-FACE DATASETS - CAMERA VIEWPOINT AND CHOICE OF FURNITURE: in bedrooms the coarse styles control where the camera is and the middle styles select which furniture appears, which an art historian would call composition and subject matter rather than style Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[4]`	1.0	the tagger's words
claim_relations	builds-on → style transfer literature Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[0]`		the tagger's words
claim_relations	builds-on → Progressive GAN (Karras et al.) Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[1]`		the tagger's words
claim_relations	builds-on → the Progressive GAN setup of Karras et al. [30] Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[2]`		the tagger's words
claim_relations	extends → the parallel self-modulation work of Chen et al. [6] Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[3]`		the tagger's words
concept_edges	style —operationalised-as→ spatially invariant statistics Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. strength of the relation: 1 also in: style · spatially invariant statistics	`concept_edges[0]`		the tagger's words
content_tags	50,000 — images used to compute the FID metric Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`content_tags.quantities[0]`		the tagger's words
content_tags	1024×1024 — output resolution (progressive from 4x4 to 1024x1024) Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`content_tags.quantities[1]`		the tagger's words
content_tags	The generator learns, without supervision, to separate high-level attributes such as pose and identity from stochastic variation such as freckles and hair, and allows intuitive, scale-specific control of synthesis. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted.	`content_tags.findings[0]`		the tagger's words
content_tags	The new generator improves on the state of the art in traditional distribution-quality metrics, interpolates demonstrably better, and better disentangles the latent factors of variation. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted.	`content_tags.findings[1]`		the tagger's words
content_tags	FID Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE RECOVERED FROM THE PAPER'S OWN CHARACTERS. The recorded quote was LETTERS-ONLY IDENTICAL to this stretch of the staged text but not an exact substring of it - spaces, hyphens, ligature glyphs or mathematical operators differed, so it could never anchor. This is the file's own text between the same two letters: not produced by a rule about the damage, copied out of the file. Round-trip checked (the recovered span is letters-only equal to the original) and length-checked. The text as originally recorded is kept in evidence as extracted. also in: FID	`content_tags.variables[1]`		the tagger's words
content_tags	adaptive instance normalization (AdaIN) Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. also in: adaptive instance normalization (AdaIN)	`content_tags.definitions[0]`		the tagger's words
content_tags	EVIDENCE CORRECTED (RE-GROUNDED) · FFHQ Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. EVIDENCE CORRECTED: RE-GROUNDED. The published string was not verbatim in EITHER copy of the paper. A quote field promises the author's words; this held 'a new, highly…' - which welds the abstract's phrasing 2,456 characters earlier to the introduction's parenthetical. The introduction states both in one sentence and that is what is quoted now. also in: FFHQ	`content_tags.definitions[1]`		the tagger's words
content_tags	mean Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. also in: mean	`content_tags.statistics[1]`		the tagger's words
content_tags	standard-deviation Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: standard-deviation	`content_tags.statistics[2]`		the tagger's words

Interpretive choices

Tags that could not be placed in the text:

none — every tag was located in the text

What was deliberately not tagged, and why, is recorded per paper by the tagger. Where a paper has no such record, this page says so rather than implying that nothing was declined.